average-timings.sh

#!/usr/bin/env bash

grep "real" $1 \
  | grep -o "m\(.*\)s\(.*\)" \
  | sed -e 's/^m/0m/' \
  | tr -s 'ms' ' ' | awk '{ print ($1*60)+$2 }' \
  | sort -n \
  | tail -n +2 | tail -r | tail -n +2 \
  | awk '
    {
      sum += $1;
      n++;
    }
    END {
      print sum / n;
    }'

benchmark.sh

#!/usr/bin/env bash

while getopts "n:" arg; do
  case $arg in
    n) TEST_ITERATIONS="$OPTARG"; ;;
  esac
done

TEST_CMD="${@:  -1}"
TEST_DESTINATION="$(mktemp -d)"

echo "TEST_CMD: $TEST_CMD"
echo "TEST_DESTINATION: $TEST_DESTINATION"
echo "TEST_ITERATIONS: $TEST_ITERATIONS"

echo

for i in `seq 1 "$TEST_ITERATIONS"`; do
  {
    time $TEST_CMD $TEST_DESTINATION/$i;
  } 2>&1
done

dataset.csv